Type-4 Phosphodiesterase (PDE4) Blockade Reduces NETosis in Cystic Fibrosis

Abstract

Neutrophilic inflammation is a key determinant of cystic fibrosis (CF) lung disease. Neutrophil-derived free DNA, released in the form of extracellular traps (NETs), significantly correlates with impaired lung function in patients with CF, underlying their pathogenetic role in CF lung disease. Thus, specific approaches to control NETosis of neutrophils migrated into the lungs may be clinically relevant in CF. We investigated the efficacy of phosphodiesterase (PDE) type-4 inhibitors, in vitro, on NET release by neutrophils from healthy volunteers and individuals with CF, and in vivo, on NET accumulation and lung inflammation in mice infected with Pseudomonas aeruginosa. PDE4 blockade curbed endotoxin-induced NET production and preserved cellular integrity and apoptosis in neutrophils, from healthy subjects and patients with CF, challenged with endotoxin, in vitro. The pharmacological effects of PDE4 inhibitors were significantly more evident on CF neutrophils. In a mouse model of Pseudomonas aeruginosa chronic infection, aerosol treatment with roflumilast, a selective PDE4 inhibitor, gave a significant reduction in free DNA in the BALF. This was accompanied by reduced citrullination of histone H3 in neutrophils migrated into the airways. Roflumilast-treated mice showed a significant improvement in weight recovery. Our study provides the first evidence that PDE4 blockade controls NETosis in vitro and in vivo, in CF-relevant models. Since selective PDE4 inhibitors have been recently approved for the treatment of COPD and psoriasis, our present results encourage clinical trials to test the efficacy of this class of drugs in CF.

Keywords: PDE4 inhibition; cystic fibrosis; lung damage; neutrophil; neutrophil extracellular traps.

FIGURE 1

(A) Neutrophils isolated from healthy subjects were allowed to adhere on fibrinogen-coated slides in the absence or presence of endotoxin (10 μg/ml) and cultured for 4 or 18 h. At the end of the incubation, samples were fixed and stained for DNA (blue staining, DRAQ-5). (B) Neutrophils from healthy subjects pretreated with CFTRInh-172 (10 µMoles/L) or a vehicle were allowed to adhere on fibrinogen-coated surfaces and stimulated with different agonists for 4 or 18 h. At the end of the incubation, free DNA was quantitated. Briefly, 2.5 µl of endonuclease (nuclease micrococcal, from Staphylococcus aureus 50 U/ml) per 500 µl of the sample was added, and samples were incubated at 37°C for 10 min. The reaction was stopped with 5 µl of EDTA (0.5 M). Samples were centrifuged at 10.000 x g for 3 min, and supernatants were stored at -20°C until DNA quantification by Quant-iT™ dsDNA high-sensitivity assay kit (Invitrogen). Results are means ± SEM of experiments with cells from 4 different donors performed in duplicates.

FIGURE 2

Neutrophils isolated from healthy subjects were allowed to adhere on fibrinogen-coated slides in the presence of endotoxin, with or without rolipram (10 µMoles/L) and cultured for 18 h. At the end of the incubation, samples were fixed and stained for DNA (blue staining, DRAQ-5), myeloperoxidase (green staining, FITC-conjugated anti-myeloperoxidase antibody), and F-actin (red staining, TRIC-phalloidin) and analyzed by confocal microscopy. The figure shows images representative of three different experiments.

FIGURE 3

(A) Neutrophils from healthy subjects, pretreated for 2 min with increasing concentrations of RNO (0–1,000 nMoles/L), were exposed to endotoxin and allowed to adhere on fibrinogen-coated surfaces for 18 h in the presence or in the absence of CFTRinh-172 (10 µMoles/L). Unstimulated neutrophils, pretreated with increasing concentrations of RNO (0–1,000 nMoles/L), were incubated in parallel. At the end of the incubation, free DNA was quantitated. Results are mean ± SEM of experiments performed with cells from 9–11 different donors in duplicates. *p < 0.05 (ANOVA, Dunnett test), RNO-treated vs untreated samples; **p < 0.05 (Student’s t-test), CFTRinh-172–treated vs untreated samples. (B) Neutrophils from individuals with CF, pretreated with increasing concentrations of RNO (0–1,000 nMoles/L), were incubated with endotoxin and allowed to adhere on fibrinogen-coated surfaces for 18 h. Unstimulated neutrophils, pretreated with increasing concentrations of RNO (0–1,000 nMoles/L), were incubated in parallel. Results are mean ± SEM of experiments performed with cells from seven different patients with CF (see Table 1 for patients’ characteristics). *p < 0.05 (ANOVA, Dunnett test) vs untreated samples. The presence of citrullinated histone H3 in neutrophils from three healthy donors (C) or two people with CF (D) was analyzed after 18 h of incubation. Samples were then subjected to Western blot analysis using a monoclonal antibody which specifically recognizes citrullinated histone H3. Representative Western blots are shown. The right image in panel C reports a densitometric analysis from n = 3. *p < 0.05 (Student’s t-test).

FIGURE 4

Neutrophils from healthy subjects pretreated with a vehicle (A,C,E,G) or RNO (100 nMoles/L) (B,D,F,H) were exposed to endotoxin and allowed to adhere on fibrinogen-coated surfaces for 18 h in the absence (C,D) or in the presence (G,H) of CFTRinh-172 (10 µMoles/L). Unstimulated neutrophils were incubated in parallel in the absence (A,B) or in the presence (E,F) of CFTRinh-172 (10 µMoles/L). At the end of the incubation, samples were stained for intracellular myeloperoxidase and analyzed by flow cytometry. Intact neutrophils were identified on the basis of typical SSC and FSC and analyzed for myeloperoxidase content. Results are from one representative experiment MV (neutrophil microvesicles).

FIGURE 5

Neutrophils isolated from healthy subjects (n = 16) (A,C) or from volunteers with CF (n = 7) (B,D) were treated with RNO (100 nMoles/L) (shaded bars) or a vehicle (white bars) and allowed to adhere on fibrinogen-coated surfaces in the presence or absence of CFTRinh-172 (10 μMoles/L), for 18 h with or without endotoxin. At the end of the incubation, the percentage of intact cells (A,B) and annexin V binding related to intact cells (C,D) was evaluated by flow cytometry (Figure 4). Intact neutrophils were identified by typical SSC and FSC. *p < 0.05 (Student’s t test) vs vehicle-treated samples.

FIGURE 6

C57BL/6 male mice (8–10 weeks of age) were infected i.t. with 1 × 10⁶ CFUs of MDR-RP73 embedded in agar beads and per aerosol with roflumilast (5 mg/kg) or placebo (4,4% DMSO in saline) once a day starting from 4 h after infection. Animals were killed after 28 h or 5 days of infection, and the BALF was collected. Total cells (A), neutrophils (B), and macrophages (C) were counted in the BALF. Panels show box plots of cell numbers at the time of killing(28 h and 5 days after infection) of vehicle- (n = 9 per group) or roflumilast-treated mice (n = 9 killed at 28 h and n = 8 killed at 5 days after infection). The horizontal lines mark the median of values, the edges of each box mark the 25th and 75th percentiles, and the vertical lines indicate the highest and lowest values, respectively, which are not outliers (values greater than 1.5 times the length of the box were considered outliers and excluded from the analysis). *p < 0.05 (ANOVA, Dunnett’s test) vs vehicle-treated mice.

FIGURE 7

C57BL/6 male mice were treated as described in Figure 6 and killed after 28 h or 5 days of infection. The BALF was collected for measurement of free DNA in supernatants and citrullinated histone H3 in supernatants and in cell lysates. (A) shows box plots of free-DNA values at the time of killing (28 h and 5 days after infection) of vehicle- (n = 9 per group) or roflumilast-treated mice (n = 9 killed at 28 h and n = 8 killed at 5 days after infection). *p < 0.05 (ANOVA, Dunnett’s test) vs vehicle-treated mice. The presence of citrullinated histone H3 was analyzed by Western blotting in BALF supernatants (B) as well as in lysates of BALF cells from mice killed at 28 h and 5 days after infection (C). Pools of BALF supernatants and of cell lysates from all animals of each group were subjected to Western blotting using a monoclonal antibody which recognizes mouse with citrullinated histone H3.

FIGURE 8

C57BL/6 male mice were treated as described in Figure 6. (A). Body weight of mice was monitored daily before treatment to evaluate the health status. *p < 0.05 (ANOVA, Dunnett test) vs vehicle-treated mice. (B) shows box plots of weight loss values at the time of killing(28 h and 5 days after infection) of vehicle- or roflumilast-treated mice. *p < 0.05 (Student’s t-test) vs vehicle-treated mice.